Leukocytes in the regulation of pain and analgesia

Transcription

1 Leukocytes in the regulation of pain and analgesia H. L. Rittner, 1 H. Machelska, and C. Stein Klinik für Anästhesiologie und operative Intensivmedizin, Charité-Universitätsmedizin Berlin, Campus Benjamin Franklin, Germany Abstract: When tissue is destroyed or invaded by leukocytes in inflammation, numerous mediators are delivered by the circulation and/or liberated from resident and immigrated cells at the site. Proalgesic mediators include proinflammatory cytokines, chemokines, protons, nerve growth factor, and prostaglandins, which are produced by invading leukocytes or by resident cells. Less well known is that analgesic mediators, which counteract pain, are also produced in inflamed tissues. These include anti-inflammatory cytokines and opioid peptides. Interactions between leukocyte-derived opioid peptides and opioid receptors can lead to potent, clinically relevant inhibition of pain (analgesia). Opioid receptors are present on peripheral endings of sensory neurons. Opioid peptides are synthesized in circulating leukocytes, which migrate to inflamed tissues directed by chemokines and adhesion molecules. Under stressful conditions or in response to releasing agents (e.g., corticotropin-releasing factor, cytokines, noradrenaline), leukocytes can secrete opioids. They activate peripheral opioid receptors and produce analgesia by inhibiting the excitability of sensory nerves and/or the release of excitatory neuropeptides. This review presents discoveries that led to the concepts of pain generation by mediators secreted from leukocytes and of analgesia by immune-derived opioids. J. Leukoc. Biol. 78: ; Key Words: inflammation neuropathy hyperalgesia cytokines chemokines opioid peptides INTRODUCTION In inflammation, numerous mediators are delivered by the circulation (e.g., bradykinin) and/or liberated at the site by resident or migrated leukocytes. Although these mediators contribute to the body s ability to counteract the destruction of tissue integrity, they also elicit pain by activation of specialized primary afferent neurons (PAN), nociceptors. They are defined as receptors preferentially sensitive to a noxious stimulus or to a stimulus which would become noxious if prolonged (definition of the International Association for the Study of Pain, ref. [1]). Trigeminal and dorsal root ganglia (DRG) contain nociceptor cell bodies, which give rise to myelinated A and unmyelinated C fibers. A and C fibers transduce and propagate noxious stimuli to the dorsal horn of the spinal cord from where these stimuli are transmitted to the brain. At the level of the spinal cord and at supraspinal sites, various neurotransmitters come into play, which together with environmental and cognitive factors, contribute to the eventual sensation of pain. Strictly speaking, analgesia is defined as the inhibition of pain in man, and antinociception is defined as the inhibition of behavioral responses to noxious stimuli in animals. Although central mechanisms also play a prominent role, the following review will focus on the peripheral injured tissue itself. PRO-ALGESIC MECHANISMS Inflammatory pain is characterized by an increased response to mechanical and heat stimuli, which are normally, mildly painful (mechanical or thermal hyperalgesia) [1]. After tissue injury, mediators are delivered by the circulation (e.g., bradykinin), and local tissue macrophages and dendritic cells are activated. The inflammatory response is amplified by migration of leukocytes into the inflamed tissue, by production of inflammatory mediators, including cytokines and nerve growth factor (NGF), as well as tissue acidification. Carrageenan, complete Freund s adjuvant (CFA), lipopolysaccharide (LPS), zymosan, or glycogen is used experimentally to induce nonspecific inflammation in animals. Neuropathic pain can arise following injury of peripheral nerves, when damaged or neighboring, undamaged nerve fibers are sensitized or fire ectopically. It is also characterized by mechanical and thermal hyperalgesia. In addition, patients and animals with neuropathy are sensitive to stimuli that do not evoke a pain behavior under normal conditions, e.g., touching, cooling, or warming the affected site (allodynia) [1]. The most common animal models of neuropathic pain are tight ligation of spinal nerves and tight or loose ligation of the sciatic nerve [2]. Proinflammatory cytokines Tumor necrosis factor (TNF- ) TNF- is the prototypic proinflammatory cytokine as a result of its role in initiating a cascade of cytokines and growth factors in the inflammatory response. TNF- exerts its effects through 1 Correspondence: Klinik für Anästhesiologie und operative Intensivmedizin, Campus Benjamin Franklin, Charité-Universitätsmedizin Berlin, Hindenburgdamm 30, D Berlin, Germany. heike.rittner@charite.de Received June 2, 2005; accepted July 11, 2005; doi: /jlb /05/ Society for Leukocyte Biology Journal of Leukocyte Biology Volume 78, December

2 two known receptors, TNFR1 and TNFR2, which are expressed on inflammatory cells and in the DRG and are up-regulated in inflammation or neuropathy [3, 4]. TNFR expression on peripheral nerve terminals has not been examined but seems likely because of its expression in DRG and axons. In rodents, the intraplantar injection of TNF- into noninflamed paws produces a short-lived, dose-dependent, mechanical hyperalgesia via indirect mechanisms (prostaglandins, sympathetic amines, or NGF) [5, 6], as well as direct sensitization. A direct action on nociceptors was postulated in electrophysiological studies, when TNF- was applied onto the sciatic nerve or injected subcutaneously (s.c.) [7]. Antisera against TNF- can block hyperalgesia induced by intraplantar carrageenan or CFA injection [5, 6]. A role for TNF- in inflammatory pain has also been confirmed using TNFR1 knockout (KO) mice. The intensity of hyperalgesia after intraplantar carrageenan injection was significantly lower in these animals. Monocytes and tissue macrophages are the primary cell sources for TNF- synthesis in response to a wide variety of agents including viruses, bacterial, and parasitic products, complement, cytokines, ischemia, and trauma (Fig. 1) [5, 8]. TNF- initiates a cascade activating IL-6 and IL-8, then IL-1 and NGF, and finally, prostaglandins or sympathetic amines. Treatment of animals with neuropathic injury using anti- TNF- antibodies and especially the combination of anti- TNF- and anti-il-1 antibodies significantly reduced signs of neuropathic pain [9]. In vitro perfusion of DRG cells with TNF- elicits neuronal discharges in A and C fibers, which are markedly higher and longer lasting after nerve injury [10]. This demonstrates an increased sensitivity of injured neurons to TNF-, where TNF- -induced hyperalgesia in this model is mediated via TNFR1 [4]. IL-1 Hyperalgesic effects of IL-1 have first been demonstrated by injection into hindpaws of rats without inflammation or nerve injury (reviewed in ref. [5]). Several mechanisms have been proposed to mediate this indirect action, including prostaglandins, nitric oxide, NGF, and bradykinin. However, there is also evidence of a direct action of IL-1 on nociceptors: Intraplantar injection of IL-1 potentiates action potentials in response to thermal or mechanical stimuli in rat DRG neurons and decreases the mechanical threshold for nerve firing [11]. In the in vitro skin preparation, brief exposure to IL-1 facilitates the heat-induced release of a calcitonin gene-related peptide (a proalgesic mediator) from peptidergic sensory neurons [12, 13]. Only one of the IL-1R subtypes, IL-1RI, is expressed in DRG cells [13]. However, IL-1RI localization has not been shown on peripheral nerve terminals. Using autoradiography, we did not detect IL-1 binding on nerves in inflamed or noninflamed paws but on leukocytes in inflamed paws [14]. Therefore, IL-1RI expression in the periphery requires further studies. In hindpaw inflammation induced by carrageenan or LPS, IL-1Ra can reduce, but not completely block, hyperalgesia in rats and in mice [5]. The same effect can be seen in CFAinduced hindpaw inflammation or in mice after intraperitoneal injection of acetic acid [5, 15]. It is interesting that our group Fig. 1. Pro- and analgesic mechanisms in inflammation. In early inflammation, an inflammatory agent stimulates the migration of leukocytes, e.g., granulocytes (G) and monocytes (M), into the inflamed tissue. Here, these leukocytes as well as resident cells initiate a cascade of cytokines including TNF- and interleukins (ILs), chemokines [CXC chemokine ligand 8 (CXCL8), CXCL1], NGF, and secondary mediators, such as sympathetic amines, leukotriene B 4 (LTB 4 ), and prostaglandins, culminating in hyperalgesia. TNF-, IL-6, IL-1, and bradykinin can also have direct hyperalgesic effects on nociceptors. During ongoining, late inflammation, lymphocytes (L) and monocytes/macrophages (M) start to produce anti-inflammatory cytokines, such as IL-4, IL-10, IL-13, and IL-1 receptor antagonist (IL-1ra). These cytokines inhibit the proinflammatory cytokines, such as TNF-, IL-1, and IL-6, and block the cascade. has shown analgesic effects of IL-1 via release of opioid peptides from leukocytes in CFA-induced inflammation (see below). In neuropathic pain models, blocking of IL-1R reduces thermal hyperalgesia, mechanical allodynia, and immunoreactivity for TNF- in mice [16]. However, the mechanisms remain unclear. IL-6 Intraplantar injection of IL-6 induces hyperalgesia in normal rats, presumably via prostaglandins [5] (Fig. 1). Direct effects have been proposed by other groups: In the rat, in vitro skin 1216 Journal of Leukocyte Biology Volume 78, December

3 preparation basal IL-6 did not induce heat sensitization when applied alone but was effective in the presence of soluble IL-6R (sil-6r) [12, 17], which is needed for cells that express only the signaling receptor unit gp130. The complex of sil-6r and IL-6 can then induce a signal to activate cells. This so-called trans-signaling expands the spectrum of cells responsive to IL-6 to include neurons. Hyper-IL-6 (HIL-6), a fusion protein of IL-6 and sil-6r, was designed to mimic the effects of the IL-6-sIL-6R complex. In vitro exposure of DRG neurons to HIL-6 potentiated heat-activated, inward currents dependent on Janus tyrosine kinase and protein kinase C [18]. As seen for IL-1 and TNF-, endogenous IL-6 is also important for the development of hyperalgesia: Antisera-neutralizing IL-6 inhibited LPS-induced hyperalgesia in rats [5]. IL-6 KO mice have reduced thermal and mechanical hyperalgesia after injection of carrageenan into the hindpaws [19]. In neuropathic mice, the IL-6 mrna level was up-regulated in DRG and correlated well with the development of nerve injury-induced thermal hyperalgesia and mechanical allodynia [20]. IL-6 KO mice had reduced thermal and mechanical hyperalgesia in the sciatic nerve loose ligation model [19]. In summary, TNF-, IL-1, and IL-6 have hyperalgesic properties as exogenous and endogenous mediators in inflammatory and neuropathic pain. Their effects seem to be mediated indirectly via prostaglandins and sympathomimetic amines or directly, by sensitizing sensory neurons. Chemokines Several observations indicated a role of chemokines in hyperalgesia. Chemokine receptors such as CXC chemokine receptor 4 (CXCR4) and CC chemokine receptor 4 (CCR4) are expressed on subpopulations of DRG neurons, and their corresponding ligands can induce calcium influx [21]. However, in neonates, the percentage of neurons responding to chemokine stimulation was significantly lower than the percentage responding to the well-known pronociceptive compounds bradykinin or capsaicin. Direct intraplantar injection of some chemokines [CC chemokine ligand 5 (CCL5), CXCL12, and CCL22] induces hyperalgesia in normal animals [21]. In addition, certain chemokines (human CXCL8 or rat CXCL1) can indirectly cause hyperalgesia through release of sympathetic amines when applied s.c. into the hindpaw [5, 8]. In contrast, our own unpublished data suggest that intraplantar injection of CXCL1 as well as CXCL2/3 in normal rats does not evoke thermal or mechanical hyperalgesia for up to 12 h post-injection, despite significant recruitment of granulocytes (H. L. Rittner, unpublished results). Such differences could be explained by the doses or the behavioral tests used. In a recent study, a connection between the CCR1 ligand CCL3 (macrophage inflammatory protein-1 ) and the capsaicin receptor transient receptor potential vanilloid 1 (TRPV1) was shown in DRG neurons [22]. Activation of CCR1 resulted in increases in TRPV1-mediated Ca 2 influx and increased sensitivity of TRPV1 to its ligand capsaicin. In vivo, the latency in the mouse hot-plate test was reduced after local injection of CCL3. Thus, CCL3 seems to be capable of enhancing the sensitivity of TRPV1 through a G protein-dependent signaling pathway. In carrageenan-induced inflammation, human CXCL8 and rat CXCL1 were shown to contribute to hyperalgesia in rats and mice [5, 8]. These observations are in contrast to our own studies in CFA-induced inflammation. In this model, concomitant injection of CXCL2/3 and CFA induced a threefold increase in the number of neutrophils in the inflamed paw after 2 h, and mechanical and thermal hyperalgesia were not changed [23]. Furthermore, the local inhibition of CXCL1 and CXCL2/3 in inflamed paws had no effect on mechanical hyperalgesia, despite significant reduction in the number of infiltrating neutrophils after 2 h [24] (see below). The differences between our studies and those by Cunha et al. [5] might be related to the model of inflammation (CFA vs. carrageenan) or to the type of behavioral test used. Although we determined the amount of mechanical pressure required for the rat to withdraw its paw (modified Randall-Selitto method), Cunha et al. [5] used a substantially more painful stimulus, which induced sympathetic activation with a freezing reaction and apnea. In a neuropathic pain model, disruption of the gene for CCR2 (attracting monocytes) prevented mechanical allodynia and decreased the number of monocytes [25]. In summary, the contribution of chemokines to hyperalgesia may depend on their type, the model of pain, the time-frame, and the dose of chemokine used. Some of the indirect effects are mediated by the ability of chemokines to attract monocytes. NGF NGF belongs to the family of neurotrophin proteins and governs the innervation of target tissues during development. It also plays an important role in neuronal survival and maintenance of connectivity. Besides its action on neurons, NGF also has immunomodulatory effects. After binding to its high-affinity receptor tyrosine receptor kinase A (trka) on peripheral terminals of sensory neurons, NGF is internalized and retrogradely transported to the somata in the DRG cells. NGF is a potent regulator of gene expression of neuropeptides, such as the calcitonin gene-related peptide and substance P, receptors such as TRPV1, and ion channels such as tetrodotoxin-resistant sodium channels. The role of endogenous NGF has mostly been examined using proteins that block the bioactivity of NGF, as deletion of NGF or trka in mice is lethal. Injection of NGF into the normal rat paw induces a longlasting thermal as well as mechanical hyperalgesia [26]. The detection of sensitizing properties of NGF has been corroborated by the analysis of transgenic mice overexpressing NGF in the epidermis [27]. These animals developed thermal hyperalgesia without any signs of inflammation of the skin. Blocking of NGF by trka-immunoglobulin G or anti-ngf antibodies in inflammation prevents the development of thermal hyperalgesia and the sensitization of nociceptors, despite normal development of inflammation, as measured by tissue edema [26, 28]. Thus, NGF seems to be part of the inflammatory cascade initiated by inflammatory cytokines such as IL-1 and TNF- (Fig. 1) [5]. NGF also influences neuropathic pain; however, the data seem to be controversial. Treatment with anti-ngf in the sciatic nerve loose ligation model reduced hyperalgesia, but the effect showed a delayed onset, short duration, and depen- Rittner et al. Leukocytes, pain, and analgesia 1217

4 dency on the dose [29]. NGF blockade seems to produce analgesia, also in other pain states, including bone cancer pain [30]. In summary, neutralization of NGF produces analgesia in most pain models studied. ANALGESIC MECHANISMS Leukocytes are the source, not only of hyperalgesic but also of analgesic mediators. Among the best-characterized and clinically relevant systems are the endogenous opioid peptides and receptors. Other analgesic mediators include anti-inflammatory cytokines as well as somatostatin and the endocannabinoids. Anti-inflammatory cytokines In later stages of inflammation, cytokines, which limit inflammation and counteract hyperalgesia, are produced (references in ref. [5]). Pretreatment with IL-4, IL-10, and IL-13 dosedependently blocked hyperalgesia induced by carrageenan, bradykinin, and TNF- but did not affect hyperalgesia induced by CXCL8 and prostaglandins (Fig. 1). Furthermore, the endogenous role of anti-inflammatory cytokines to limit hyperalgesia, induced by carrageenan, bradykinin, and TNF-, was demonstrated by application of antisera against IL-4, IL-10, and IL-13 (references in ref. [5]). Endogenous sources of IL-4 and IL-13 to produce analgesia are mast cells and lymphocytes, respectively. Analgesic actions of IL-4, IL-10, and IL-13 are not only seen in paw inflammation but also in a model of peritonitis and knee-joint incapacitation induced by zymosan (references in ref. [5]). In summary, during ongoing inflammation, analgesic cytokines counteract the effects of the proinflammatory hyperalgesic cytokines generated in the early stages. Immune-derived opioid peptides Opioid receptors Three cdnas and their genes have been identified, encoding the -, -, and -opioid receptors [31]. All three receptors can mediate pain inhibition, and they are found throughout the central and peripheral nervous system. Recent interest has focused on the characterization of opioid receptors on nociceptors, as their activation can inhibit pain directly at its origin without unwanted central side-effects. Peripheral opioid receptors are synthesized in DRG and are intra-axonally transported to the peripheral nerve endings (Fig. 2). Opioid receptors belong to seven-transmembrane domain G-protein-coupled receptors. Upon activation by opioid ligands, they couple to inhibitory G-proteins (G i/o ), which leads to inhibition of calcium and/or sodium channels and to a decreased level of neuronal cyclic adenosine monophosphate. Consistent with these effects, opioids attenuate the excitability of nociceptors, the propagation of action potentials, and the release of excitatory proinflammatory neuropeptides (substance P, calcitonin gene-related peptide) from central and peripheral nociceptor endings [32]. All of these mechanisms result in analgesia. Inflammation of peripheral tissue leads to increased synthesis and axonal transport of opioid receptors in DRG neurons, Fig. 2. Migration of opioid-producing leukocytes and opioid secretion within inflamed tissue. P-selectin, intercellular adhesion molecule-1 (ICAM-1), and platelet-endothelial cell adhesion molecule-1 (PECAM-1) are up-regulated on vascular endothelium of blood vessels. L-selectin and integrins 4 and 2 are coexpressed by opioid peptide-containing, circulating leukocytes. These cells also coexpress receptors for chemokines, which are presented on endothelium. L- and P-selectin mediate rolling of opioid-containing cells along the vascular endothelium. These cells can then be activated by chemokines, which upregulate integrins. 4 and 2 integrins and ICAM-1 mediate opioid-containing leukocyte adhesion (firm attachment) to and diapedesis through the endothelium. In these processes, adhesion molecules interact with their respective ligands (other adhesion molecules) expressed on leukocytes and endothelium. Although PECAM-1 is present, it does not seem to contribute to the migration of opioid-containing cells. Once extravasated and accumulated in inflamed extravascular tissues, leukocytes can be stimulated by stress (e.g., swim stress, surgery) or injection of releasing agents such as corticotropin-releasing factor (CRF), IL-1, and noradrenaline (NA) to secrete opioid peptides. CRF, IL-1, and NA [derived from postganglionic sympathetic neurons (PGSN)] elicit opioid release by activating CRF receptors (CRFR), IL-1Rs, and adrenergic receptors (AR) on leukocytes, respectively. Opioids bind to peripheral opioid receptors (produced in DRG and transported to peripheral endings of nociceptors, i.e., PAN), and by the inhibition of substance P release, they produce analgesia. resulting in their up-regulation and enhanced G-protein coupling at peripheral nerve terminals [33 36]. Also, the number of nociceptor endings increases, and the perineural barrier is disrupted, which facilitates the access of opioid agonists to their receptors. All these effects lead to enhanced analgesic efficacy of opioids at their peripheral receptors in inflammation [32]. Opioid peptide production in leukocytes Three families of opioid peptides are well-characterized: the endorphins, enkephalins, and dynorphins. They bind to all three opioid receptors with varying affinities. Each family derives from a distinct gene and the respective precursors, i.e., proopiomelanocortin (POMC), proenkephalin, and prodynorphin. Additional, selective -ligands, the endomorphins, have been isolated, but their precursors are not known yet [37]. All opioid peptides are found in leukocytes, but endorphins deriving from POMC have been studied most extensively. POMC 1218 Journal of Leukocyte Biology Volume 78, December

5 processing occurs in the endoplasmic reticulum and the trans- Golgi network. The enzymatic machinery required for this process includes carboxypeptidase E, the prohormone convertases PC1 and PC2, and the binding protein 7B2 [38]. Recently, we detected -endorphin, POMC, and all processing enzymes in leukocytes in the blood and within inflamed tissue in rats [38]. Thus, leukocytes can process POMC into functionally active -endorphin. Furthermore, Met-enkephalin, dynorphin, and endomorphins are also detectable in leukocytes of inflamed tissue. The opioid-containing cells are T- and B-lymphocytes, granulocytes, and monocytes/macrophages [34, 39 41]. Thus, opioid peptides are processed and present in the circulation and in cells infiltrating injured tissue. Migration of opioid-containing leukocytes to inflamed tissue The recruitment of leukocytes from the circulation into inflammatory sites involves a well-orchestrated set of events. This begins with rolling along the endothelial cell wall mediated predominantly by L-, P-, and E-selectins. Then, leukocytes are activated by chemokines released from inflammatory cells and presented on the endothelium. This leads to the up-regulation and increased avidity of integrins, which mediate the firm adhesion of leukocytes to endothelial cells via, e.g., ICAM-1. Finally, leukocytes transmigrate through the endothelium mediated by, e.g., PECAM-1 [42]. These events are also involved in the endogenous control of inflammatory pain (Fig. 2). In the CFA model, we have shown that integrin 4, CXCL1, and CXCL2 are expressed by leukocytes and that P- and E-selectins, ICAM-1, and PECAM-1 are up-regulated on endothelium [24, 43 45]. Expression of CXCL1 and CXCL2 mrnas and protein contents significantly increased during the course of inflammation [23, 24]. It is important that L-selectin, integrin 2, and CXCR2 are coexpressed by opioid-containing leukocytes. Furthermore, pretreatment of rats with a selectin blocker (fucoidin) or with antibodies against ICAM-1, 4, 2, CXCL1, or CXCL2 substantially decreases the number of opioid-containing leukocytes accumulating in the inflamed tissue [24, 44 46] and in consequence, abolishes endogenous peripheral opioid analgesia (see below). This suggests that circulating opioid-producing leukocytes are directed to inflamed tissue by these adhesion molecules and chemokines and eventually secrete the opioids to inhibit pain [41, 47]. Release of opioid peptides from leukocytes Regulated secretion of peptides requires secretory granules deriving from the Golgi network. As discussed above, leukocytes express the entire machinery required for POMC processing into functionally active -endorphin [38]. Furthermore, in macrophages, monocytes, granulocytes, and lymphocytes, -endorphin is present in secretory granules arranged at the cell periphery, ready for the exocytosis. As in the pituitary, CRF and IL-1 release opioid peptides from leukocytes in vitro in the CFA inflammatory model. These effects are specific to CRF- and IL-1 Rs (coexpressed by opioid-containing leukocytes), and they are calcium-dependent and are mimicked by elevated extracellular concentrations of potassium [14, 41, 47 50]. This is consistent with a regulated pathway of release from secretory vesicles, as in neurons and endocrine cells. Furthermore, NA can release -endorphin from leukocytes in an AR-specific manner in vitro [51]. Adrenergic 1, 2, and to a lesser degree, 2 receptors are expressed on -endorphincontaining inflammatory cells located in close proximity to sympathetic nerve fibers in inflamed paws, and chemical ablation of these fibers abolishes intrinsic opioid analgesia. In summary, CRF, IL-1, and sympathetic neuron-derived NA can act on their respective receptors on leukocytes to release opioid peptides (Fig. 2). Analgesia produced by immune-derived opioid peptides Analgesic effects of CRF, cytokines, and NA. In our model, CRF and IL-1 injected into inflamed paws produce CRF-, IL-1 -, and opioid receptor-specific analgesia. -Endorphin plays a major role, but Met-enkephalin and dynorphin are also involved [24, 48, 50, 52]. Immunosuppression with cyclosporine A, depletion of granulocytes, and blockade of selectins, ICAM-1, CXCL1, and CXCL2 decrease the number of opioid-containing cells and the efficacy of CRF- or IL-1 induced analgesia, demonstrating that immigrating leukocytes are the target for CRF and IL-1 [24, 46, 48, 53]. These results are in contrast to hyperalgesia induced by IL-1, IL-1, IL-6, and TNF- [5, 6], as mentioned above. However, they are in line with other studies about local analgesic effects of exogenous CRF [54, 55], IL-6, and TNF- [56] in inflammatory pain and of endogenous CRF and IL-1 in electroacupuncture-induced analgesia in inflammation [57]. Furthermore, intraplantar NA activates 1, 2, and 2 AR on inflammatory cells, resulting in antinociception mediated by -endorphin and neuronal opioid receptors in our model. Important parameters, which may explain differences between studies, include the absence or presence, the duration, and the model of inflammation. For example, in all our studies, the presence of opioid peptide-containing inflammatory cells and the coexpression of the respective stimulatory receptors on such cells were prerequisites. Together, CRF, IL-1, and NA can act at their respective receptors on leukocytes to release opioid peptides, which subsequently bind to neuronal opioid receptors, leading to pain relief (Fig. 2). Endogenous opioid analgesia. Stress is a natural stimulus triggering inhibition of pain [58]. In rats with CFA inflammation, stress induced by cold water (4 C) swim (for 1 min) elicits potent analgesia only in inflamed paws. At early inflammatory stages (6 h), all three families of opioid peptides and opioid receptors are involved, and at later stages (4 6 days), -endorphin, acting at and receptors, dominates. Whereas at early stages, peripheral and central opioid receptors contribute, at later stages, endogenous analgesia is mediated exclusively by peripheral opioid receptors [52, 59, 60]. Thus, peripheral opioid mechanisms of pain control become more prevalent with the duration and severity of inflammation. Endogenous triggers of this stress-induced analgesia are locally produced CRF and sympathetic neuron-derived NA, as this effect is abolished by local neutralization of CRF and by sympathetic blockade [49, 51, 52]. Leukocytes are the source of opioids, as immunosuppression with cyclosporine A or whole body irradiation and depletion of monocytes/macrophages also Rittner et al. Leukocytes, pain, and analgesia 1219

6 block the stress effect [39, 59, 61]. Moreover, this analgesia is extinguished by inhibiting the extravasation of -endorphincontaining leukocytes by blockade of L- and P-selectins, 4 and 2 integrins, or of ICAM-1, as mentioned above [44 46] (Fig. 2). Examining factors that increase homing of opioidcontaining cells to injured tissue, we showed that hematopoietic growth factors strongly mobilized granulocytes in the blood but resulted only in a minor increase in the number of opioidcontaining leukocytes in inflamed paws and in no change of CRF- or swim stress-induced analgesia [62]. Increasing the migration of opioid-containing cells to inflamed tissue with local injections of CXCL2 did not result in stronger analgesia either. Most probably, this was a result of the small number of neuronal opioid receptors at this early (2 h) stage of inflammation [23]. It is important that intrinsic analgesia, increasing with the duration of inflammation (2 h 4 days), coincides with an increase in the number of opioid-containing leukocytes, in the number of opioid receptors, and in the efficacy of opioid receptor G-protein coupling in peripheral sensory neurons [33, 34, 40]. CLINICAL IMPLICATIONS AND PERSPECTIVES Myriad pronociceptive mediators are involved in inflammatory as well as neuropathic pain. Peripherally acting drugs undoubtedly have advantages, such as optimized drug concentrations at the site of injury while avoiding systemically active drug levels, adverse systemic effects, or drug interactions. Selective or nonselective inhibitors of cyclooxygenase (COX) are most frequently used as peripherally acting pain medications, as they inhibit the final common pathway of most proalgesic mediators, the production of prostaglandins. However, these drugs have serious side-effects, such as gastrointestinal ulcer formation, bleeding, and/or thromboembolic events. Therefore, new treatments are under current investigation [63]. Cytokine antagonists, originally developed for treatment of autoimmune diseases such as rheumatoid arthritis, have been shown to also produce analgesia in patients [64]. Opioid receptors on peripheral nerve terminals can mediate analgesia in patients with chronic arthritis but also in bone pain or postoperative pain after dental, laparoscopic, urinary bladder, and knee surgery [32, 65]. Opioid peptides, such as -endorphin and Met-enkephalin, were found in human synovial lining cells, mast cells, lymphocytes, and macrophages [65, 66]. Blocking intra-articular opioid receptors (with their antagonist naloxone) significantly increased postoperative pain, demonstrating tonic release of endogenous opioid peptides from leukocytes in a stressful (e.g., postoperative) situation [66]. In contrast to the central nervous system, it appears that immune cell-derived opioids do not readily produce cross-tolerance to morphine at peripheral opioid receptors, as intra-articular morphine is an equally potent analgesic in patients with and without opioid-producing, inflammatory synovial cells [65]. Thus, it may be interesting to explore the opioid production/ release and the migration of opioid-containing leukocytes as possible treatment options, not only in conventional but also in complementary and alternative (e.g., accupuncture [57]) medicine. Furthermore, the important role of certain adhesion molecules and chemokines in the trafficking of opioid-containing cells to injured tissues indicates that antiadhesion or antichemokine strategies for the treatment of inflammatory diseases may in fact carry a significant risk to exacerbate pain. It would be highly desirable to identify stimulating factors and strategies that selectively attract opioid-producing cells and increase peripheral opioid receptor numbers in damaged tissue. It is important that opioid analgesia, resulting from the described neuroimmune interactions, occurs in peripheral tissues and therefore, is devoid of central opioid side-effects (such as depression of breathing, nausea, clouding of consciousness, addiction, and tolerance) and of typical side-effects produced by COX inhibitors. ACKNOWLEDGMENTS This work was supported by Deutsche Forschungsgemeinschaft (KFO 100/1), Bundesministerium für Bildung und Forschung (#01GZ0311), and European Society of Anesthesiology. We thank Christine Voigts for the preparation of the illustrations. REFERENCES 1. Merskey, H., Bogduk, N. (1994) Pain terminology. In Classification of Chronic Pain. IASP Task Force on Taxonomy (H. Merskey, N. Bogduk, eds.), Seattle, WA, IASP, Bennet, G. J. (1994) Animal models of neuropathic pain. In Proceedings of the 7th World Congress on Pain, Progress in Pain Research and Management (G. F. Gebhart, D. L. Hammond, T. S. Jensen, eds.), Seattle, WA, IASP, Li, Y., Ji, A., Weihe, E., Schafer, M. K. (2004) Cell-specific expression and lipopolysaccharide-induced regulation of tumor necrosis factor (TNF ) and TNF receptors in rat dorsal root ganglion. J. Neurosci. 24, Schäfers, M., Sorkin, L. S., Geis, C., Shubayev, V. I. (2003) Spinal nerve ligation induces transient upregulation of tumor necrosis factor receptors 1 and 2 in injured and adjacent uninjured dorsal root ganglia in the rat. Neurosci. Lett. 347, Cunha, F. Q., Ferreira, S. H. (2003) Peripheral hyperalgesic cytokines. Adv. Exp. Med. Biol. 521, Woolf, C. J., Allchorne, A., Safieh-Garabedian, B., Poole, S. (1997) Cytokines, nerve growth factor and inflammatory hyperalgesia: the contribution of tumor necrosis factor. Br. J. Pharmacol. 121, Junger, H., Sorkin, L. S. (2000) Nociceptive and inflammatory effects of subcutaneous TNF. Pain 85, Cunha, T. M., Verri Jr., W. A., Silva, J. S., Poole, S., Cunha, F. Q., Ferreira, S. H. (2005) A cascade of cytokines mediates mechanical inflammatory hypernociception in mice. Proc. Natl. Acad. Sci. USA 102, Schäfers, M., Brinkhoff, J., Neukirchen, S., Marziniak, M., Sommer, C. (2001) Combined epineurial therapy with neutralizing antibodies to tumor necrosis factor- and interleukin-1 receptor has an additive effect in reducing neuropathic pain in mice. Neurosci. Lett. 310, Schäfers, M., Lee, D. H., Brors, D., Yaksh, T. L., Sorkin, L. S. (2003) Increased sensitivity of injured and adjacent uninjured rat primary sensory neurons to exogenous tumor necrosis factor- after spinal nerve ligation. J. Neurosci. 23, Fukuoka, H., Kawatani, M., Hisamitsu, T., Takeshige, C. (1994) Cutaneous hyperalgesia induced by peripheral injection of interleukin-1 in the rat. Brain Res. 657, Opree, A., Kress, M. (2000) Involvement of the proinflammatory cytokines tumor necrosis factor-, IL-1, and IL-6 but not IL-8 in the development of heat hyperalgesia: effects on heat-evoked calcitonin gene-related peptide release from rat skin. J. Neurosci. 20, Journal of Leukocyte Biology Volume 78, December

7 13. Obreja, O., Rathee, P. K., Lips, K. S., Distler, C., Kress, M. (2002) IL-1 potentiates heat-activated currents in rat sensory neurons: involvement of IL-1RI, tyrosine kinase, and protein kinase C. FASEB J. 16, Mousa, S. A., Schafer, M., Mitchell, W. M., Hassan, A. H., Stein, C. (1996) Local upregulation of corticotropin-releasing hormone and interleukin-1 receptors in rats with painful hindlimb inflammation. Eur. J. Pharmacol. 311, Safieh-Garabedian, B., Poole, S., Allchorne, A., Winter, J., Woolf, C. J. (1995) Contribution of interleukin-1 to the inflammation-induced increase in nerve growth factor levels and inflammatory hyperalgesia. Br. J. Pharmacol. 115, Sommer, C., Petrausch, S., Lindenlaub, T., Toyka, K. V. (1999) Neutralizing antibodies to interleukin 1-receptor reduce pain-associated behavior in mice with experimental neuropathy. Neurosci. Lett. 270, Obreja, O., Schmelz, M., Poole, S., Kress, M. (2002) Interleukin-6 in combination with its soluble IL-6 receptor sensitizes rat skin nociceptors to heat, in vivo. Pain 96, Obreja, O., Biasio, W., Andratsch, M., Lips, K. S., Rathee, P. K., Ludwig, A., Rose-John, S., Kress, M. (2005) Fast modulation of heat-activated ionic current by proinflammatory interleukin 6 in rat sensory neurons. Brain 128, Xu, X. J., Hao, J. X., Andell-Jonsson, S., Poli, V., Bartfai, T., Wiesenfeld- Hallin, Z. (1997) Nociceptive responses in interleukin-6-deficient mice to peripheral inflammation and peripheral nerve section. Cytokine 9, Murphy, P. G., Ramer, M. S., Borthwick, L., Gauldie, J., Richardson, P. M., Bisby, M. A. (1999) Endogenous interleukin-6 contributes to hypersensitivity to cutaneous stimuli and changes in neuropeptides associated with chronic nerve constriction in mice. Eur. J. Neurosci. 11, Oh, S. B., Tran, P. B., Gillard, S. E., Hurley, R. W., Hammond, D. L., Miller, R. J. (2001) Chemokines and glycoprotein120 produce pain hypersensitivity by directly exciting primary nociceptive neurons. J. Neurosci. 21, Zhang, N., Inan, S., Cowan, A., Sun, R., Wang, J. M., Rogers, T. J., Caterina, M., Oppenheim, J. J. (2005) A proinflammatory chemokine, CCL3, sensitizes the heat- and capsaicin-gated ion channel TRPV1. Proc. Natl. Acad. Sci. USA 102, Brack, A., Rittner, H. L., Machelska, H., Shaqura, M., Mousa, S. A., Labuz, D., Zöllner, C., Schäfer, M., Stein, C. (2004) Endogenous peripheral antinociception in early inflammation is not limited by the number of opioid-containing leukocytes but by opioid receptor expression. Pain 108, Brack, A., Rittner, H. L., Machelska, H., Shaqura, M., Mousa, S. A., Labuz, D., Zöllner, C., Schäfer, M., Stein, C. (2004) Control of inflammatory pain by chemokine-mediated recruitment of opioid-containing polymorphonuclear cells. Pain 112, Abbadie, C., Lindia, J. A., Cumiskey, A. M., Peterson, L. B., Mudgett, J. S., Bayne, E. K., DeMartino, J. A., MacIntyre, D. E., Forrest, M. J. (2003) Impaired neuropathic pain responses in mice lacking the chemokine receptor CCR2. Proc. Natl. Acad. Sci. USA 100, Lewin, G. R., Rueff, A., Mendell, L. M. (1994) Peripheral and central mechanisms of NGF-induced hyperalgesia. Eur. J. Neurosci. 6, Stucky, C. L., Koltzenburg, M., Schneider, M., Engle, M. G., Albers, K. M., Davis, B. M. (1999) Overexpression of nerve growth factor in skin selectively affects the survival and functional properties of nociceptors. J. Neurosci. 19, McMahon, S. B., Bennett, D. L., Priestley, J. V., Shelton, D. L. (1995) The biological effects of endogenous nerve growth factor on adult sensory neurons revealed by a trka-igg fusion molecule. Nat. Med. 1, Ro, L. S., Chen, S. T., Tang, L. M., Jacobs, J. M. (1999) Effect of NGF and anti-ngf on neuropathic pain in rats following chronic constriction injury of the sciatic nerve. Pain 79, Sevcik, M. A., Ghilardi, J. R., Peters, C. M., Lindsay, T. H., Halvorson, K. G., Jonas, B. M., Kubota, K., Kuskowski, M. A., Boustany, L., Shelton, D. L., Mantyh, P. W. (2005) Anti-NGF therapy profoundly reduces bone cancer pain and the accompanying increase in markers of peripheral and central sensitization. Pain 115, Kieffer, B. L., Gaveriaux-Ruff, C. (2002) Exploring the opioid system by gene knockout. Prog. Neurobiol. 66, Stein, C., Schafer, M., Machelska, H. (2003) Attacking pain at its source: new perspectives on opioids. Nat. Med. 9, Zöllner, C., Shaqura, M. A., Bopaiah, C. P., Mousa, S. A., Stein, C., Schäfer, M. (2003) Painful inflammation-induced increase in opioid receptor binding and G-protein coupling in primary afferent neurons. Mol. Pharmacol. 64, Mousa, S. A., Zhang, Q., Sitte, N., Ji, R., Stein, C. (2001) -Endorphincontaining memory-cells and -opioid receptors undergo transport to peripheral inflamed tissue. J. Neuroimmunol. 115, Jeanjean, A. P., Moussaoui, S. M., Maloteaux, J. M., Laduron, P. M. (1995) Interleukin-1 induces long-term increase of axonally transported opiate receptors and substance P. Neuroscience 68, Ji, R. R., Zhang, Q., Law, P. Y., Low, H. H., Elde, R., Hokfelt, T. (1995) Expression of -, -, and -opioid receptor-like immunoreactivities in rat dorsal root ganglia after carrageenan-induced inflammation. J. Neurosci. 15, Akil, H., Owens, C., Gutstein, H., Taylor, L., Curran, E., Watson, S. (1998) Endogenous opioids: overview and current issues. Drug Alcohol Depend. 51, Mousa, S. A., Shakibaei, M., Sitte, N., Schafer, M., Stein, C. (2004) Subcellular pathways of -endorphin synthesis, processing, and release from immunocytes in inflammatory pain. Endocrinology 145, Przewlocki, R., Hassan, A. H., Lason, W., Epplen, C., Herz, A., Stein, C. (1992) Gene expression and localization of opioid peptides in immune cells of inflamed tissue: functional role in antinociception. Neuroscience 48, Rittner, H. L., Brack, A., Machelska, H., Mousa, S. A., Bauer, M., Schäfer, M., Stein, C. (2001) Opioid peptide-expressing leukocytes: identification, recruitment, and simultaneously increasing inhibition of inflammatory pain. Anesthesiology 95, Cabot, P. J., Carter, L., Gaiddon, C., Zhang, Q., Schäfer, M., Loeffler, J. P., Stein, C. (1997) Immune cell-derived -endorphin. Production, release, and control of inflammatory pain in rats. J. Clin. Invest. 100, von Andrian, U. H., Mackay, C. R. (2000) T-cell function and migration. Two sides of the same coin. N. Engl. J. Med. 343, Mousa, S. A., Machelska, H., Schafer, M., Stein, C. (2000) Co-expression of -endorphin with adhesion molecules in a model of inflammatory pain. J. Neuroimmunol. 108, Machelska, H., Brack, A., Mousa, S.A., Schopohl, J.K., Rittner, H.L., Schäfer, M., Stein, C. (2004) Selectins and integrins but not plateletendothelial cell adhesion molecule-1 regulate opioid inhibition of inflammatory pain. Br. J. Pharmacol. 142, Machelska, H., Mousa, S. A., Brack, A., Schopohl, J. K., Rittner, H. L., Schäfer, M., Stein, C. (2002) Opioid control of inflammatory pain regulated by intercellular adhesion molecule-1. J. Neurosci. 22, Machelska, H., Cabot, P. J., Mousa, S. A., Zhang, Q., Stein, C. (1998) Pain control in inflammation governed by selectins. Nat. Med. 4, Cabot, P. J., Carter, L., Schäfer, M., Stein, C. (2001) Methionine-enkephalin- and Dynorphin A-release from immune cells and control of inflammatory pain. Pain 93, Schäfer, M., Carter, L., Stein, C. (1994) Interleukin 1 and corticotropinreleasing factor inhibit pain by releasing opioids from immune cells in inflamed tissue. Proc. Natl. Acad. Sci. USA 91, Schäfer, M., Mousa, S. A., Zhang, Q., Carter, L., Stein, C. (1996) Expression of corticotropin-releasing factor in inflamed tissue is required for intrinsic peripheral opioid analgesia. Proc. Natl. Acad. Sci. USA 93, Mousa, S. A., Bopaiah, C. P., Stein, C., Schäfer, M. (2003) Involvement of corticotropin-releasing hormone receptor subtypes 1 and 2 in peripheral opioid-mediated inhibition of inflammatory pain. Pain 106, Binder, W., Mousa, S. A., Sitte, N., Kaiser, M., Stein, C., Schafer, M. (2004) Sympathetic activation triggers endogenous opioid release and analgesia within peripheral inflamed tissue. Eur. J. Neurosci. 20, Machelska, H., Schopohl, J. K., Moussa, S. A., Schäfer, M., Stein, C. (2003) Different mechanisms of intrinsic pain inhibition in early and late inflammation. J. Neuroimmunol. 141, Machelska, H., Stein, C. (2002) Immune mechanisms in pain control. Anesth. Analg. 95, Hargreaves, K. M., Dubner, R., Costello, A. H. (1989) Corticotropin releasing factor (CRF) has a peripheral site of action for antinociception. Eur. J. Pharmacol. 170, McLoon, L. K., Sandnas, A. M., Nockleby, K. J., Wirtschafter, J. D. (2002) Reduction in vesicant-induced cellular inflammation and hyperalgesia by local injection of corticotrophin-releasing factor in rabbit eyelid. Inflamm. Res. 51, Czlonkowski, A., Stein, C., Herz, A. (1993) Peripheral mechanisms of opioid antinociception in inflammation: involvement of cytokines. Eur. J. Pharmacol. 242, Sekido, R., Ishimaru, K., Sakita, M. (2004) Corticotropin-releasing factor and interleukin-1 are involved in the electroacupuncture-induced anal- Rittner et al. Leukocytes, pain, and analgesia 1221

8 gesic effect on inflammatory pain elicited by carrageenan. Am. J. Chin. Med. 32, Terman, G. W., Shavit, Y., Lewis, J. W., Cannon, J. T., Liebeskind, J. C. (1984) Intrinsic mechanisms of pain inhibition: activation by stress. Science 226, Stein, C., Hassan, A. H., Przewlocki, R., Gramsch, C., Peter, K., Herz, A. (1990) Opioids from immunocytes interact with receptors on sensory nerves to inhibit nociception in inflammation. Proc. Natl. Acad. Sci. USA 87, Stein, C., Gramsch, C., Herz, A. (1990) Intrinsic mechanisms of antinociception in inflammation: local opioid receptors and -endorphin. J. Neurosci. 10, Brack, A., Labuz, D., Schiltz, A., Rittner, H. L., Machelska, H., Schäfer, M., Reszka, R., Stein, C. (2004) Tissue monocytes/macrophages in inflammation: hyperalgesia versus opioid-mediated peripheral antinociception. Anesthesiology 101, Brack, A., Rittner, H. L., Machelska, H., Beschmann, K., Sitte, N., Schäfer, M., Stein, C. (2004) Mobilization of opioid-containing polymorphonuclear cells by hematopoetic growth factors and influence on inflammatory pain. Anesthesiology 100, Marceau, F., Regoli, D. (2004) Bradykinin receptor ligands: therapeutic perspectives. Nat. Rev. Drug Discov. 3, Cole, P., Rabasseda, X. (2004) The soluble tumor necrosis factor receptor etanercept: a new strategy for the treatment of autoimmune rheumatic disease. Drugs Today (Barc) 40, Stein, C., Pfluger, M., Yassouridis, A., Hoelzl, J., Lehrberger, K., Welte, C., Hassan, A. H. (1996) No tolerance to peripheral morphine analgesia in presence of opioid expression in inflamed synovia. J. Clin. Invest. 98, Stein, C., Hassan, A. H., Lehrberger, K., Giefing, J., Yassouridis, A. (1993) Local analgesic effect of endogenous opioid peptides. Lancet 342, Journal of Leukocyte Biology Volume 78, December